Wild-Card Quick Reads
by Vincent Verhei
The wild-card round saw four quarterbacks make their playoff debuts. As a group, Marcus Mariota, Jared Goff, Tyrod Taylor, and Blake Bortles completed 53 percent of their passes, averaging 5.2 yards per throw, with four touchdowns and two interceptions.
The wild-card round also saw four veteran quarterbacks with ample playoff experience; Alex Smith, Cam Newton, Matt Ryan, and Drew Brees respectively made their seventh, seventh, ninth, and 12th playoff starts. As a group, they completed 69 percent of their passes and averaged 8.9 yards per throw.
In terms of wins and losses, one debuting quarterback beat another when Bortles' Jaguars beat Taylor's Bills, while one veteran beat another when Brees' Saints beat Newton's Panthers. We had two matchups between a vet and a newbie, and the results were split -- Ryan's Falcons beat Goff's Rams, but Mariota's Titans beat Smith's Panthers. This had us all asking: does playoff experience really matter?
First things first: wins are a team stat, of course, not an individual stat. And measuring teams' experience levels simply by what their quarterback has done is inherently faulty. Yes, Drew Brees has oodles of playoff experience, but Sunday's game against Carolina was the first postseason action for Michael Thomas, Alvin Kamara, Ryan Ramczyk, Marshon Lattimore, Kenny Vaccaro, Marcus, Williams, etc. More often than not, though, when quarterbacks make their playoff debuts, it's the first time in a while their teams have made the playoffs too. (Right, Bills fans?)
With that in mind, let's start breaking down some data. From the 2000 season through 2016, in matchups between quarterbacks with different levels of experience, the more seasoned quarterback has gone 98-78 (.588). Of course this data is faulty too -- the quarterbacks with the most playoff experience tend to be the best players on the best teams. In the 17 years we studied, the quarterback with the most playoff games was Tom Brady with 34. He was followed by Peyton Manning (26), Ben Roethlisberger (20), Aaron Rodgers, and Donovan McNabb (both with 16). It's no surprise those men and their teams have winning records against the A.J. McCarrons and Joe Webbs of the world.
We can somewhat control for this my looking at more specific data. We might expect a quarterback in his first playoff game to lose to one in his tenth, but what happens against an opponent in his second game? Or third or fourth?
The following table looks at exactly that question.
| Playoff Records by Quarterback Experience, 2000-2016 | ||||||||||
| Playoff Start |
Opponent Playoff Start | With More Experience |
With Less Experience |
|||||||
| 1 | 2 | 3 | 4 | 5+ | Total | |||||
| 1 | 5-5 (.500) |
5-3 (.625) |
2-2 (.500) |
3-5 (.375) |
7-22 (.241) |
22-37 (.373) |
-- | 17-32 (.347) |
||
| 2 | 3-5 (.375) |
1-1 (.500) |
2-3 (.400) |
4-8 (.333) |
4-10 (.286) |
14-27 (.342) |
3-5 (.375) |
10-21 (.323) |
||
| 3 | 2-2 (.500) |
3-2 (.600) |
3-3 (.500) |
2-2 (.500) |
9-8 (.529) |
19-17 (.528) |
5-4 (.556) |
11-10 (.524) |
||
| 4 | 5-3 (.625) |
8-4 (.667) |
2-2 (.500) |
3-3 (.500) |
20-13 (.606) |
38-25 (.603) |
16-10 (.615) |
21-14 (.600) |
||
| 5+ | 22-7 (.759) |
10-4 (.714) |
8-9 (.471) |
13-20 (.394) |
41-41 (.500) |
94-81 (.537) |
74-59 (.556) |
19-21 (.475) |
||
Here's how to read that table, if it's a little confusing. The left-hand column lists quarterbacks in their first, second, third, and fourth playoff starts, with all starts after that dumped to one bin in the end. From left to right, each column lists how those quarterbacks fared against opponents in their first start, second start, etc. So 10 times from 2000 to 2016, quarterbacks have made playoff debuts against each other. Obviously, one has to win and one has to lose each game, so they are .500 in those games. Eight games have pitted a player in his first start against one in his second; the debutants have gone 5-3 in those games. You can get the same data from the other viewpoint by reading down the columns instead -- second-game quarterbacks have gone 3-5 against players in their first starts.
The most interesting data is in the columns on the right-hand side, those listing the totals. Here we see that quarterbacks have gone 22-37 (.373) overall in their first playoff games, but remember that includes a 5-5 record against each other. Remove those games, and they have gone 17-32 (.347) against more experienced quarterbacks. There is also a column listing how quarterbacks have fared when they have more experience, but of course that's irrelevant for players in their postseason debuts.
It's interesting that quarterbacks in their second games have actually fared worse than those in their debuts, but there's a simple explanation for this: a lot of quarterbacks get a win in the wild-card round in their first playoff games, then are rewarded with a road game against a higher-seeded opponent in the divisional round. Often, they lose that second game, and then never make it back to the playoffs again. (See: Aaron Brooks, Matt Schaub, and in all likelihood, Brock Osweiler). Those who do make it to a third game win more often than they lose, and from that point forward win expectancy more or less levels off.
So to answer our original question, yes, playoff experience absolutely matters, at least for a quarterback's first few games.
One final question: Does this data change based on the experience gap between two players? There's a lot of noise and fluctuation in that data, but the answer appears to be "not really."
| Playoff Records Based on Difference in QB Experience, 2000-2016 |
||||
| Experience Edge |
Record | Experience Edge |
Record | |
| 1 | 13-16 (.448) |
6 | 6-5 (.545) |
|
| 2 | 11-6 (.647) |
7 | 7-2 (.778) |
|
| 3 | 9-9 (.500) |
8 | 6-3 (.667) |
|
| 4 | 7-4 (.636) |
9 | 5-6 (.455) |
|
| 5 | 5-9 (.357) |
10+ | 29-18 (.617) |
|
All that said and done, we now await the divisional round of the playoffs, which will offer us these matchups, in chronological order:
- Matt Ryan in his 10th start vs. Nick Foles in his second.
- Marcus Mariota in his second start vs. Tom Brady in his 35th.
- Blake Bortles in his second start vs. Ben Roethlisberger in his 21st.
- Drew Brees in his 13th start vs. Case Keenum in … his first.
Good luck, young men.
| Quarterbacks | |||||||||||
Rk |
Player |
Team |
CP/AT |
Yds |
TD |
INT |
Sacks |
Total DYAR |
Pass DYAR |
Rush DYAR |
Opp |
1. |
Drew Brees | NO | 24/33 |
376 |
2 |
1 |
1 |
220 |
220 |
0 |
CAR |
First-down passing: 10-of-15, 138 yards, five first downs. Second-down passing: 11-of-12, 210 yards, with every completion gaining enough yards for a first down. Third-/fourth-down passing: 2-of-5, 28 yards, two conversions, plus one sack, one interception, and one intentional grounding. That's a DVOA of 78.2%, 187.7%, and -66.3% by down. | 2. |
Cam Newton | CAR | 25/40 |
349 |
2 |
0 |
4 |
169 |
169 |
0 |
NO |
Newton played well just about everywhere -- but not scoring range. Inside the New Orleans 40, he went 6-of-16 for 57 yards with one touchdown, one intentional grounding, and two sacks. | 3. |
Matt Ryan | ATL | 21/30 |
218 |
1 |
0 |
3 |
103 |
109 |
-7 |
LARM |
Ryan killed the Rams on middle-range passes. On throws tha traveled 3 to 7 yards beyond the line of scrimmage, he went 12-of-16 for 99 yards and a touchdown. He had 10 first downs on those throws, but only three on his other 14. | 4. |
Alex Smith | KC | 24/33 |
264 |
2 |
0 |
4 |
77 |
80 |
-4 |
TEN |
Smith's last five throws of the first quarter were all complete, for gains of 45, 27, 26, 18, and 13 yards, the last a touchdown. In the second half, though, he went 5-of-10 for just 33 yards and a pair of sacks. Part of that is likely due to the Travis Kelce injury -- on the other hand, after Kelce got hurt, Smith finished the first half going 6-of-6 for 68 yards and five first downs, including a touchdown. | 5. |
Marcus Mariota | TEN | 19/31 |
205 |
2 |
1 |
2 |
39 |
8 |
12 |
KC |
Mariota's DYAR totals include 19 receiving DYAR for his touchdown catch. He also gets 24 passing DYAR on the play. No, we are not sure this is the fairest way to measure that, but it's what we went with. On third-down passes in the first half, he went 1-of-3 for 12 yards with on conversion and one sack; in the second half, he went 4-of-4 for 39 yard and four conversions, including a touchdown. He threw eight deep balls; the first seven were all incomplete, but the eighth resulted in a 22-yard touchdown. | 6. |
Jared Goff | LARM | 24/45 |
272 |
1 |
0 |
3 |
34 |
34 |
0 |
ATL |
Goff did not pick up a first down until the Rams were down 13-0 in the second quarter. In the first quarter, he went 1-of-5 for 3 yards and two sacks. He didn't have a lot of luck throwing to his left: 5-of-17 for just 37 yards, though one of those completions did result in a touchdown. | 7. |
Blake Bortles | JAC | 12/22 |
87 |
1 |
0 |
2 |
-26 |
-49 |
24 |
BUF |
Bortles did not come by that 55 percent completion rate through a bevy of Hail Marys. He only threw one deep pass all day, completing it for 20 yards. No, no, it was all those inaccurate short throws that ruined his statistical day. Bortles' average incomplete pass traveled only 3.9 yards past the line of scrimmage; every other starter's average incompletion this week was at least twice as deep. He didn't even attempt a pass as deep as 10 yards until the two-minute warning in the first half. On third downs, he went 2-of-7 for 5 yards with no conversions and one sack. (He did throw a touchdown on his only fourth-down pass.) | 8. |
Tyrod Taylor | BUF | 17/37 |
134 |
0 |
1 |
2 |
-37 |
-26 |
-11 |
JAC |
Taylor got a boost of 85 DYAR due to opponent adjustments, nearly double that of anyone else this week, and still winds up way down here. He was actually reasonably effective in the first 19 minutes of the game, but then he threw an interception, and was brutal from that point forward: 9-of-25 for 66 yards, plus a 7-yard DPI, one sack, and only two total first downs. Outside the Buffalo 40, he went 2-of-14 for 13 yards with no first downs and two sacks. |
| Five Best Running Backs by DYAR (Total) | ||||||||||||
Rk |
Player |
Team |
Runs |
Rush Yds |
Rush TD |
Rec |
Rec Yds |
Rec TD |
Total DYAR |
Rush DYAR |
Rec DYAR |
Opp |
1. |
Derrick Henry | TEN | 23 |
159 |
1 |
2/2 |
35 |
0 |
43 |
25 |
19 |
KC |
Five first downs on the ground, including four runs of 10 yards or more, while being hit for no gain or a loss just three times. He would have scored even higher but he did have a fumble. His two catches: a 6-yard gain on second-and-10 and a 29-yard gain on second-and-6. | 2. |
Christian McCaffrey | CAR | 6 |
16 |
0 |
6/8 |
101 |
1 |
41 |
-7 |
47 |
NO |
McCaffrey did not run for a first down and his longest gain was just 7 yards. However, four of his receptions produced first downs, including a pair of third-down conversions and a 56-yard touchdown. | 3. |
Mike Tolbert | BUF | 4 |
15 |
0 |
1/1 |
10 |
0 |
12 |
1 |
10 |
JAC |
Tolbert's best carry was an 11-yard gain on first-and-10. His one catch was a 10-yard gain on third-and-7. | 4. |
Alvin Kamara | NO | 10 |
23 |
1 |
1/3 |
10 |
0 |
9 |
14 |
-5 |
CAR |
Kamara's 2-yard touchdown run was his only first down on the ground all day. His longest run gained only 5 yards, and three of his runs went for no gain or a loss. His only catch was a 10-yard gain on first-and-20. | 5. |
Kareem Hunt | KC | 11 |
42 |
1 |
3/3 |
5 |
0 |
7 |
18 |
-11 |
TEN |
Hunt's best runs were a 1-yard touchdown and a 16-yard gain on first-and-10. He was hit for a loss twice. His two catches were a 7-yard gain on second-and-10 and a 2-yard loss on second-and-20. |
| Five Best Running Backs by DYAR (Rushing) | ||||||||||||
Rk |
Player |
Team |
Runs |
Rush Yds |
Rush TD |
Rec |
Rec Yds |
Rec TD |
Total DYAR |
Rush DYAR |
Rec DYAR |
Opp |
1. |
Derrick Henry | TEN | 23 |
159 |
1 |
2/2 |
35 |
0 |
43 |
25 |
19 |
KC |
2. |
Todd Gurley | LARM | 14 |
101 |
0 |
4/10 |
10 |
0 |
-20 |
23 |
-43 |
ATL |
Gurley had five first downs on the ground, including gains of 14, 26, and 33 yards. None of his receptions were successful plays; his longest was a 4-yard gain on third-and-10. | 3. |
Kareem Hunt | KC | 11 |
42 |
1 |
3/3 |
5 |
0 |
7 |
18 |
-11 |
TEN |
4. |
Alvin Kamara | NO | 10 |
23 |
1 |
1/3 |
10 |
0 |
9 |
14 |
-5 |
CAR |
5. |
Mike Tolbert | BUF | 4 |
15 |
0 |
1/1 |
10 |
0 |
12 |
1 |
10 |
JAC |
| Worst Running Back by DYAR (Total) | ||||||||||||
Rk |
Player |
Team |
Runs |
Rush Yds |
Rush TD |
Rec |
Rec Yds |
Rec TD |
Total DYAR |
Rush DYAR |
Rec DYAR |
Opp |
1. |
Leonard Fournette | JAC | 21 |
57 |
0 |
3/5 |
21 |
0 |
-54 |
-46 |
-9 |
BUF |
Fournette only ran for three first downs in his 21 carries. His second-longest run was a 10-yard gain on third-and-13, which padded his yardage total but did little for his DYAR. He was hit for no gain or a loss six times. And he lost 25 DYAR for doing it against the defense ranked 31st in DVOA against the run this season. |
| Worst Running Back by DYAR (Rushing) | ||||||||||||
Rk |
Player |
Team |
Runs |
Rush Yds |
Rush TD |
Rec |
Rec Yds |
Rec TD |
Total DYAR |
Rush DYAR |
Rec DYAR |
Opp |
1. |
Leonard Fournette | JAC | 21 |
57 |
0 |
3/5 |
21 |
0 |
-54 |
-46 |
-9 |
BUF |
| Five Best Wide Receivers and Tight Ends by DYAR | ||||||||||
Rk |
Player |
Team |
Rec |
Att |
Yds |
Avg |
TD |
Total DYAR |
Opp |
1. |
Julio Jones | ATL | 9 |
10 |
94 |
10.4 |
1 |
62 |
LARM |
Jones' DYAR totals include 52 DYAR receiving, 10 DYAR rushing for his one carry for 13 yards. Six of his catches went for first downs, including a gain of 27 and an 8-yarder. | 2. |
Michael Thomas | NO | 8 |
9 |
131 |
16.4 |
0 |
60 |
CAR |
Thomas also gained six first downs, the longest a 46-yarder. | 3. |
Demarcus Robinson | KC | 4 |
4 |
57 |
14.2 |
1 |
42 |
TEN |
All four of Robinson's catches gained 13, 14, or 15 yards and a first down, including a third-down conversion. | 4. |
Ted Ginn | NO | 4 |
6 |
115 |
28.8 |
1 |
38 |
CAR |
Ginn's 80-yard touchdown was worth 36 DYAR, so otherwise he basically had a replacement-level day. He did have two other first downs, including a third-down conversion. | 5. |
Travis Kelce | KC | 4 |
4 |
66 |
16.5 |
1 |
38 |
TEN |
14-yard gain on third-and-6; 27-yard gain on first-and-10; 13-yard touchdown; 12-yard gain on second-and-18. |
| Worst Wide Receiver or Tight End by DYAR | ||||||||||
Rk |
Player |
Team |
Rec |
Att |
Yds |
Avg |
TD |
Total DYAR |
Opp |
1. |
Nick O'Leary | BUF | 2 |
6 |
22 |
11.0 |
0 |
-16 |
JAC |
O'Leary's first two targets went for first downs, but obviously he didn't catch another pass all day. He actually loses 3 DYAR due to opponent adjustments. The Jacksonville defense does a lot of things well, but covering tight ends isn't one of them. |
Comments
35 comments, Last at 12 Jan 2018, 5:00am
#2 by ammek // Jan 08, 2018 - 7:53am
Not so for me. I'm unsure what conclusions to draw from the data (which might be the point), but happy to have them laid out and analyzed.
Playoff experience is certainly a narrative that the announcers like to fall back upon. It would be interesting to see DYAR for these QBs, and perhaps win-loss records adjusted for home-field advantage: it feels more relevant that Tom Brady keeps playing for a team with a first-round bye and HFA (ie, a better team) rather than the fact that he has however much postseason experience.
#13 by roguerouge // Jan 08, 2018 - 12:21pm
I think the only thing to draw out of it is that QBs in their first or second game have bad win-loss records, so adjust your expectations/gambling accordingly. I'm certainly interested to see how this plays out with the Vikings, for example.
#9 by jtr // Jan 08, 2018 - 10:41am
Sometimes the Quick Reads articles are interesting to me and sometimes I just skip them. It's easy to skip them without complaining that you don't like it. I'm glad they put in the effort to pull something interesting out of the numbers in this feature, otherwise it would just be a weekly spreadsheet dump.
#31 by Mountain Time … // Jan 11, 2018 - 10:11pm
Really? I didn't have any trouble, and I didn't even know there was a paragraph explaining it until after reading the chart in different directions.
Re: jtr
I think it's totally fine to provide feedback if you thought something was TLDR, laudable even. Raiderfan left some useful information for the writers, as did the rest of us who were fine with it
#5 by Aaron Brooks G… // Jan 08, 2018 - 9:18am
I would expect some of the young vs old comparison is based on team DVOA, too.
Young guys running into a peak Brady-Manning-Roethlisberger-Rodgers-McNabb team were probably at an overall talent disadvantage, too.
#4 by wiesengrund // Jan 08, 2018 - 8:31am
I also looked in playoff debuts by QBs a week ago, and found that controlling for defense DVOA might be the most telling stat of all:
Going back 15 yrs, among those to start their playoff-career 1-0 are: Osweiler, Tebow, Yates, Schaub, Smith, Cutler, Sanchez, Flacco, Garrard, Delhomme
The 0-1 Group includes: Prescott, Newton, Luck, Stafford, Dalton, Rodgers, Ryan, Romo, Rivers, Palmer, E Manning, Brees, Green
(Notice, this did not remove games were debutants were pitted against each other but the big picture remains the same.)
Simply anecdotally, you could make the case that bad QBs are more likely to win their debut, since they usually got there by a good defense and/or running game and by being asked to do less. If they are nervous in their debut, it has much less effect. Yes, Tebow lives in infamy for the OT TD, but that was his 21st attempt of the game. Most of the 1-0 crowd had great defenses behind them and had not to do it by themselves (with Alex smith being probably the biggest exception - he carried his team to a win even though his usually great defense allowed 30+ on that day).
The 1-0 crowd are mostly good QBs that had very little around them. Getting jittery there sinks the teams chances much more drastically. My hunch is therefore that Quality QBs usually have less complete teams around them, and are therefore much more susceptible to debut-losses.
Looking at it through defensive DVOA lense might illustrate the point, although there certainly are exceptions as well. Goff for instance had a pretty complete team around him on Saturday, and he simply could not pull out the Alex Smith where he could compensate for a usually strong unit (ST) failing on that day. Another exception is Cam Newton: He had a top 3 defense and the 4th rushing DVOA behind him in his debut and he still lost it to a great 49ers team.
But overall I think that point would be worth exploring a little more in depth.
#6 by Aaron Brooks G… // Jan 08, 2018 - 9:26am
Stafford put up 380 yards at 9 yards per pass and converted 7/10 3rd downs, for 143 DYAR. A pro-rated 2288 DYAR would have been in a dead-heat with Rodgers and Brady for 2nd that year (but still 300 behind Brees).
The Lions defense gave up 45 points, 34 1st downs, 626 yards, 7/11 on 3rd, 3/4 on 4th, and 359 DYAR. God wouldn't have outscored Brees that week.
Man, looking back, that Tebow game came out of nowhere. 10/21 sure, but for 316 yards with another 51 rushing, and 208 DYAR. And DVOA doesn't like mad-bomber games!
#11 by JoeyHarringtonsPiano // Jan 08, 2018 - 11:37am
In the 0-1 group, it's kind of unfair to include Palmer, because his ACL got shredded during his very first snap.
Stafford (as Aaron Brooks Good Twin mentions above), Rodgers, and Prescott played very well in their debuts, but their opponents had HOF-level quarterbacks who simply played even better.
Don't remember all the details of the other quarterbacks.
#20 by wiesengrund // Jan 09, 2018 - 4:01am
I did throw out Palmers "real" debut for that exercise. His first true start I counted was the 14-24 loss to the Jets in 2009.
In terms of Stafford, Rodgers, Prescott: Yes, that is exactly my point. Those QBs, even on a good day, simply couldn't keep up with their defense letting the other team have a HOF caliber performance.
#7 by Bob Smith // Jan 08, 2018 - 10:05am
This is one of those "rough" subjects when I reflect back on my guys (Dan Marino) career. Dan couldn't play good enough to help us (the Dolphins) win his 1st PO game, but then he plays good enough to help get us to the S.B. in his 2nd year (yes, I know we still had the nucleus of our '82 S.B. team led by the great QB David Woodley). But after '84, Dan could not play good enough to help us have any more real success in the Po's for 15 seasons. Sometimes it depends on the QB and his ability to play multiple good games in the PO's, and Dan was not one of those guys.
#8 by jtr // Jan 08, 2018 - 10:36am
> a lot of quarterbacks get a win in the wild-card round in their first playoff games, then are rewarded with a road game against a higher-seeded opponent in the divisional round. Often, they lose that second game, and then never make it back to the playoffs again. (See: Aaron Brooks, Matt Schaub, and in all likelihood, Brock Osweiler).
Tim Tebow's on the list, in a game that all Pittsburgh fans have desperately tried to forget. TJ Yates is also probably going to end up in the club, though he did play a snap or two in relief of Matt Moore in the playoffs last year without throwing a pass.
#10 by nat // Jan 08, 2018 - 11:13am
...a lot of quarterbacks get a win in the wild-card round in their first playoff games, then are rewarded with a road game against a higher-seeded opponent in the divisional round.
This makes me wonder how this analysis would work if it was based on "years of playoff experience" rather than games.
After all, one way to start a lot of playoff games is to be on a strong team than goes deep into the playoffs. You'd expect a team like that to do better in the next postseason, not because of the added experience, but because they would start the season with a better roster.
#19 by Vincent Verhei // Jan 08, 2018 - 11:48pm
He only had four plays. Our standard minimum is eight.
He had 5 rushing DYAR on one carry. (Yes, he fumbled, but it came at the end of a fourth-down conversion.)
He had -60 DYAR on three passes (one completion for a first down, one intentional grounding, one interception.
#28 by Vincent Verhei // Jan 09, 2018 - 4:33pm
Ted Ginn has played 11 NFL seasons for five different franchises, seven different head coaches, seven different offensive coordinators, and eight different primary starting quarterbacks. The only people who have ever made him look halfway decent are Cam Newton and Drew Brees, because they are very special quarterbacks.
#23 by Eddo // Jan 09, 2018 - 12:06pm
Doesn't Brees do this fairly regularly (put a one-note deep threat in the top five of DVOA)?
Kenny Stills was #3 in DVOA in 2014 and #1 in 2013. Devery Henderson was #1 in 2008. Joe Horn was #5 in 2006.
You could actually argue that Ginn was the most established of these receivers by the time Brees was throwing to him (though Stills has a case, too).
#24 by voytron // Jan 09, 2018 - 12:08pm
"Marcus Mariota in his second start vs. Tom Brady in his 35th." 35th!!! I know everyone is tired of the Pats but in 2045 when all football is played and watched within virtual reality contact lenses people are just not going to believe those numbers.
#29 by Aaron Brooks G… // Jan 10, 2018 - 10:02am
Sure they will.
In 2045, 24 teams will make the playoffs and it will consist of 5-6 rounds.
In the Brady era, there have been half as many total playoff games played as were played in NFL history to that date.
In 2000, 9 teams had played 35 playoff games. In 2017, 15 had, and the Titans will make 16 this week.
#35 by Vincent Verhei // Jan 12, 2018 - 5:00am
There were 48 plays this year that were worth 43 or more passing DYAR. Most was a Derek Carr-Amari Cooper 87-yard touchdown on third-and-22 against the Chargers in Week 17. That was worth 52 DYAR. The second-most valuable play was a 66-yard touchdown from Marcus Mariota to Derrick Henry on second-and-14 against Jacksonville in ... Week 17.